Phase 3, open-label, randomized study comparing 3-monthly with monthly goserelin in pre-menopausal women with estrogen receptor-positive advanced breast cancer

This was a Phase 3, open-label, randomized, parallel group, multicenter trial designed to evaluate whether 3-monthly goserelin 10.8 mg is non-inferior to monthly goserelin 3.6 mg in pre-menopausal women with ER-positive advanced breast cancer by assessment of progression-free survival (PFS) rate at 24 weeks (ClinicalTrials.gov identifier: NCT01073865).

Eligible patients were randomized 1:1 to receive a goserelin 10.8 mg subcutaneous (sc) depot injection every 12 weeks (±7 days; 3-monthly goserelin 10.8 mg treatment group) or a goserelin 3.6 mg sc depot injection every 4 weeks (±7 days; monthly goserelin 3.6 mg treatment group). All patients also received a daily oral dose of tamoxifen (20 mg) (Fig. 1). The study population comprised pre-menopausal women aged ≥20 years with histologically/cytologically confirmed, hormone-sensitive breast cancer; ≥1 lesion that can be accurately assessed at baseline and is suitable for repeated assessment by CT, MRI, or plain X-ray according to Response Evaluation Criteria In Solid Tumors (RECIST) Version 1.1 [16]; and a World Health Organization performance status of 0, 1, or 2. Pre-menopausal status was defined as experiencing menses within 1 year prior to randomization, and serum concentrations for E2 ≥10 pg/mL and follicle-stimulating hormone (FSH) ≤30 mIU/mL within 4 weeks prior to randomization (for patients who had undergone a hysterectomy, only the latter criterion was required).

Patients were excluded from the study if they had received: hormonal therapies within the previous 24 weeks (or at any point for treatment of advanced breast cancer); LHRH agonist therapy within the previous 48 weeks; radiotherapy or trastuzumab/lapatinib treatment for early breast cancer within the previous 4 weeks (or at any point for treatment of advanced breast cancer); chemotherapy within the previous 4 weeks or for advanced breast cancer (except first-line chemotherapy with the presence of menses and no evidence of disease progression). Additional exclusion criteria included: relapse during or within 48 weeks of adjuvant hormonal therapy, or during or within 24 weeks of adjuvant chemotherapy; previous non-breast cancer malignancy (other than controlled basal/squamous carcinoma or cancer of the cervix); life-threatening metastatic disease; abnormal laboratory tests results; a history of disseminated intravascular coagulation or long-term anticoagulant therapy (other than antiplatelet or warfarin). Patients could not participate if they were likely to be hypersensitive to study treatments, unwilling/unable to stop taking drugs known to affect sex hormone status, were pregnant/breastfeeding, had received an unapproved drug in the previous 12 weeks, were likely to have a survival of <24 weeks, were unlikely to comply with study requirements, or had any concomitant disease that would place them at risk or may confound the results of the study. All patients provided written informed consent.

Patients continued study treatment for 24 weeks or until any of the criteria for discontinuation were met (including voluntary discontinuation, safety issues, medication non-compliance, pregnancy, disease progression, or death). Final assessment of efficacy was completed at the time of disease progression or at week 24, whichever was sooner.

The study protocol was approved by the relevant ethics committees and institutional review boards and was conducted in accordance with the Declaration of Helsinki and the International Conference on Harmonisation/Good Clinical Practice, the applicable local regulatory requirements at each study center, and the AstraZeneca policy on Bioethics. All patients provided written informed consent.

The primary endpoint was PFS rate at 24 weeks, used to assess the non-inferiority of 3-monthly goserelin 10.8 mg versus monthly goserelin 3.6 mg. Secondary efficacy endpoints included objective response rate (ORR) and serum E2 concentrations. Efficacy assessments were performed at weeks 12 and 24.

PFS rate, defined as the proportion of patients who were progression-free, was analyzed at 24 weeks. Analysis of ORR, defined as the proportion of patients with a best response of complete or partial response, was determined at each site by RECIST criteria, and assessed in patients who had measurable disease at baseline (e.g., patients with bone metastases). A sub-analysis of efficacy endpoints was carried out to determine if there were differences in treatment response between Japanese and non-Japanese patients. Pharmacodynamic effects were evaluated by recording serum E2 in both treatment groups at baseline and at 12 and 24 weeks. A central laboratory service (Covance Central Laboratory Services) performed chemiluminescent enzyme immunoassay to determine E2 (ADVIA centaur estradiol-6 III, Siemens Medical Solutions Diagnostics, NY, USA) and FSH (Access hFSH, Beckman Coulter Inc., CA, USA) concentrations. The lower limit of quantification (LLOQ) was 18.07 pg/mL for E2, and 0.6 mIU/mL for FSH.

Assessments of adverse events (AEs) and serious AEs (SAEs) were carried out to compare the safety and tolerability profiles of 3-monthly goserelin 10.8 mg and monthly goserelin 3.6 mg. Final assessment of safety was carried out at week 24 (if treatment was discontinued prior to week 24), 12 weeks after the final dose of goserelin 10.8 mg, or 4 weeks after the final dose of goserelin 3.6 mg. The safety analysis was conducted in all patients who received ≥1 treatment after randomization. A sub-analysis of the incidence of AEs/SAEs was also conducted in Japanese and non-Japanese patients. SAEs were classified by the Medical Dictionary for Regulatory Activities preferred term and system organ class.

Based on a meta-analysis by Klijn et al., the proportion of patients who had not progressed during the first 24 weeks of treatment with the combination of goserelin 3.6 mg and tamoxifen was estimated to be approximately 70 % [9]. Therefore, a sample size of 216 patients (108 per group) was planned to demonstrate non-inferiority based on a pre-specified margin of −17.5 % between treatment groups (deemed clinically acceptable) using a 2-sided 95 % confidence interval (CI) with 80 % power. The non-inferiority margin was chosen based on an admissible 75 % relative efficacy of the 3-monthly goserelin 10.8 mg treatment regimen versus the monthly goserelin 3.6 mg treatment regimen.

Two-sided 95 % CIs were calculated using the score based method recommended by Newcombe and Altman [17]. For the primary endpoint, non-inferiority of the 3-monthly goserelin 10.8 mg treatment regimen versus the monthly goserelin 3.6 mg treatment regimen was to be concluded if the lower limit of the 95 % CI for the difference was greater than −17.5 %. This non-inferiority limit was only applicable to the primary endpoint of PFS at 24 weeks in the overall population; no non-inferiority testing was carried out for other endpoints or subgroups. ORRs at 24 weeks were determined for each treatment, and the difference and 95 % CI numerically compared. Mean serum E2 concentrations were compared using analysis of covariance (ANCOVA), with treatment group, baseline serum E2 concentrations, and ethnicity (Japanese versus non-Japanese) as covariates. Clinical laboratory data were summarized by treatment group using descriptive statistics.

This study was conducted at 58 centers: India (15 centers), Japan (19 centers), Korea (4 centers), the Philippines (7 centers), Thailand (8 centers), and Taiwan (5 centers). In total, 222 patients were randomized to receive 3-monthly goserelin 10.8 mg (n = 109) or monthly goserelin 3.6 mg (n = 113). Of these patients, 30.6 % (n = 68) discontinued from the study [25.7 % (n = 28) in the goserelin 10.8 mg group; 35.4 % (n = 40) in the goserelin 3.6 mg group] with the most common reason for withdrawal being disease progression (Fig. 2).

All patients were ER-positive, and the majority of patients were both ER- and PgR-positive [82.6 % (n = 90) in the goserelin 10.8 mg group; 77.9 % (n = 88) in the goserelin 3.6 mg group]. The majority of patients were human epidermal growth factor receptor 2 negative [68.8 % (n = 65) in the goserelin 10.8 mg group; 74.3 % (n = 84) in the goserelin 3.6 mg group]. Patient demographics and baseline characteristics were well balanced between the treatment groups and also between the treatment groups in the Japanese and non-Japanese subgroups (Table 1).

In total, 61.5 % (n = 67) of patients in the goserelin 10.8 mg group and 60.2 % (n = 68) of patients in the goserelin 3.6 mg group were progression-free at 24 weeks (treatment difference: 1.3; 95 % CI −11.4, 13.9). Since the lower 95 % CI was above the pre-defined margin of −17.5 %, 3-monthly goserelin 10.8 mg met the criteria for non-inferiority compared with monthly goserelin 3.6 mg (Fig. 3).

In the Japanese population, 72.4 % (n = 21/29) of patients in the goserelin 10.8 mg group and 80.0 % (n = 24/30) of patients in the goserelin 3.6 mg group were progression-free at 24 weeks (treatment difference: −7.6; 95 % CI −28.5, 14.0). In the non-Japanese population, 57.5 % (n = 46/80) of patients in the goserelin 10.8 mg group and 53.0 % (n = 44/83) of patients in the goserelin 3.6 mg group were progression-free at 24 weeks (treatment difference: 4.5; 95 % CI −10.6, 19.4).

At 24 weeks, the ORR in patients with measurable disease at baseline was 23.9 % (n = 21/88) in the goserelin 10.8 mg group and 26.9 % (n = 25/93) in the goserelin 3.6 mg group (treatment difference: −3.0 %; 95 % CI −15.5, 9.7; Table 2). In Japanese patients, ORR was 25.9 % (n = 7/27) in the goserelin 10.8 mg group and 37.0 % (n = 10/27) in the goserelin 3.6 mg group (treatment difference: −11.1 %; 95 % CI −33.8, 13.2). In non-Japanese patients the ORR was 23.0 % (n = 14/61) in the goserelin 10.8 mg group and 22.7 % (n = 15/66) in the goserelin 3.6 mg group (treatment difference: 0.2 %; 95 % CI −14.2, 14.9).

Mean (standard deviation) serum E2 concentrations throughout the study period are presented in Fig. 4. At week 24, serum E2 concentrations were 20.3 (12.3) pg/mL and 24.8 (28.1) pg/mL in the goserelin 10.8 mg and goserelin 3.6 mg groups, respectively.

A total of 143 of 221 (64.7 %) patients reported ≥1 AE [65.7 % (n = 71) of patients in the goserelin 10.8 mg group and 63.7 % of patients (n = 72) in the goserelin 3.6 mg group]. The most common AEs are shown in Table 3; the majority of AEs were mild or moderate (Grade 2 or lower).

A total of 12 patients reported a SAE (4 patients in the goserelin 10.8 mg group and 8 patients in the goserelin 3.6 mg group). The incidence of SAEs was 3.7 % in the goserelin 10.8 mg group and 7.1 % in the goserelin 3.6 mg group. A total of 7 patients died in the study (4 patients in goserelin 10.8 mg group, 3 patients in goserelin 3.6 mg group). Three patients in the goserelin 10.8 mg group and 2 patients in the goserelin 3.6 mg group died due to disease progression. One patient in each group died following an AE. One patient was a 23-year-old female who had breast lumpectomy approximately 4 months prior to receiving goserelin 10.8 mg and had received zoledronic acid due to bone metastasis with no other prior medications or radiotherapy. This patient had been judged to have stable disease 2 months after their first depot 10.8 mg goserelin dose; 2 months after the second dose this patient reported chest pain, diarrhea and vomiting, and died. The other patient was a 34-year-old female who had previously had modified radical mastectomy of the right breast and had received adjuvant tamoxifen and chemotherapy (5-fluorouracil, epirubicin and cyclophosphamide) approximately 2.5 years prior to receiving goserelin. Approximately 3 months after the first depot goserelin 3.6 mg dose, this patient had been judged to have partial disease response. Approximately 2 weeks after the 6th dose, the patient developed dyspnea and pyrexia and died due to dyspnea. Neither of these deaths were judged by the investigators to be related to study treatment.

In Japanese patients, the incidence of AEs was 96.6 % (n = 28/29) in the goserelin 10.8 mg group and 83.3 % (n = 25/30) in the goserelin 3.6 mg group. Two SAEs occurred among Japanese patients, both in the goserelin 3.6 mg group. In non-Japanese patients, the incidence of AEs was 54.4 % (n = 43/79) in the goserelin 10.8 mg group and 56.6 % (n = 47/83) in the goserelin 3.6 mg group. SAEs were experienced by 5.1 % (n = 4/79) of non-Japanese patients in the goserelin 10.8 mg group and 7.2 % (n = 6/83) in the goserelin 3.6 mg group.